The performance and display quality of a thermal imaging system is often improved by normalizing or DC restoring (DCR) the output of the detectors of the thermal detector array using some DCR thermal reference source. These DCR sources have typically been active temperature sources driven to track the average temperature of a scene, although passive (non-temperature controlled) sources have been used in some applications.
The specific problem to which the invention can be applied is to implement, in a thermal imaging system, a passive DCR source by providing scene-average radiation to each thermal image detector. Preferably, such a passive DCR source would have the following optical characteristics: (a) use of the actual imaged scene as the DCR scene, (b) use of a common input aperture for both imaging and DCR, and (c) presentation of the same scene-average temperature to each detector.
Thermal imaging systems using active DCR sources, such as thermal electric coolers (TECs), are disadvantageous in a number of respects. They require drive electronics and heatsinks, and are limited in performance by response time and dynamic range. They also suffer reliability problems when they must be controlled over a wide temperature range in short periods of time.
Passive DCR sources have been used to avoid the disadvantages associated with active DCR sources. These passive DCR sources are non-temperature controlled, providing each detector with a thermal reference, such as by using scene defocusing techniques to present a local scene-average to each detector.
Thermal imaging systems may be staring systems or scanned systems in which the detector array is scanned across a scene. In staring systems, DCR is interleaved with image reading operations. In scanned systems, DCR is usually performed during the inactive portion of the scanning cycle.
This inactive period (typically on the order of 20 percent of the field rate) begins and ends at the edges of the scanned field, and includes the turnaround period and (for one-way scanned systems) the scanner retrace period. Typically, DCR is implemented by overscanning the image field to some degree.
Using an overscan period for the DCR process impacts the configuration of the DCR optics, as well as integration into the thermal imaging system optics. In particular, injecting a DCR thermal reference source into the overscan optical path very close to an intermediate image, where the optical bundle for each detector is smallest, reduces the amount of overscan required, and therefore reduces the negative impact the DCR process has on scan efficiency.
FIG. 1 illustrates a conventional scanned thermal imaging system with an active DCR source, in this case a TEC. The TEC is positioned at a DCR aperture stop that, during the overscan period, is imaged onto the scanner, the same location that the system aperture stop is imaged.
That is, during the overscan period, the optical path is broken and the detector array sees the TEC through a DCR pupil (the image of the DCR stop on the scanner), rather than the scene through the system pupil. The optical effect of imaging a pupil onto a scene is to, in effect, create an amount of defocus such that the optical energy from the scene through the pupil constitutes an average of all scene radiation (i.e., all thermal image information is lost, and the uniform scene-average temperature is provided). When the images of the two aperture stops (the system aperture stop and the DCR aperture stop) are designed to appear the same size for each detector, each detector in the array sees the entire DCR source, permitting proper DCR restoration.
Accordingly, a specific need exists for an optical scheme for implementing a passive DCR source that provides a uniform DCR thermal reference to each of the detectors in a thermal imaging system. A more general need exists for an optical design for imaging a pupil at a desired location, which could be used as such a passive DCR source.